Summary of Work: This project has explored the mechanism(s) by which E. coli MutS and MutL block recombination between closely related but non-identical DNA. At present studies have primarily focused on the role of MutS and its associated ATPase activity. In replication fidelity MutS initiates repair by binding the mismatch. The exact function of MutL is not known. Current models suggest MutL adds to the MutS-mismatch complex in an ATP-dependent fashion. MutS contains both the Walker A & B consensus sequences, and thus placing it in a superfamily of ATPases that utilize nucleotide triphosphates to drive cellular processes. The exact role of ATP binding/hydrolysis in methyl-directed repair is, however, not known but has stimulated great interest because of the internal conservation of these motifs in higher cells. We now know that nucleotide binding by E.coli, yeast, and human MutS results in loss of mismatch recognition. Consequently it is highly probable that NB (nucleotide binding) and or hydrolysis is the driving force behind MutS-dependent DNA translocation along the DNA helix contour. At present there are two models that differ in how MutS utilizes the energy from ATP to signal or activate MutH-dependent cleavage at a hemi-methylated GATC sequence. Recently, Fishel has proposed that MutS only needs to undergo NB to promote MutH incision. Unlike the long-standing model of Modrich, which proposes bidirectional loop formation via ATP turnover, Fishel suggests simple linear diffusion (or sliding clamp) by MutS along the helix contour. Indeed, based upon the Fishel model MutS mimicks K-ras GTPases where NB results in a conformational change or molecular switch. Consequently, ATP hydrolysis is the by-product of mismatch repair that allows MutS to turnover. To help clarify the mechanism behind MutS ATPase activity we wanted to address both models biochemically. Employing strategies along that used for other ATPase of this superfamily we specifically wanted to uncouple NB from hydrolysis. In this class of ATPases containing both Walker A and B motifs there is usually an active site general base that is responsible for activating a H2O molecule to carryout in-line attack on the g-PO4. We wanted to determine this site for E. coli MutS. Using alignment pile-ups we examined a series of highly conserved Glu and Asp residues among all the MutS homologs. At present there are two that appear to be interesting as it relates to ATPase activity and the ability to promote MutH depedent nicking or mismatch repair. These sites correspond to residues 553 and 661. We have evidence that suggests aspartic acid 661 is involved in hydrolysis, but is not the catalytic site residue. Indeed, changing this residue to Asn did not abolish MutS ATpase activity. Kinetic parameters, kcat and Km were only marginally affected as compared to wild-type with values of 7.4 microM min-1 and of 22 microM, respectively. MutSD66N did, however, block mismatch repair and MutH-dependent enhanced nicking. Consistent with the biochemistry studies we see through E. coli gene replacement MutSD661N is a mutator and rivals that of MutS501. Preliminary evidence with gel-retardation with strepavidin blocked substrate shows MutSD661N in not the active site carboxylate. Further studies with other mutants are in progress to unequivocally establish the catalytic site for hydrolysis. In examining further roles of mismatch repair in genome stability we have shown Escherichia coli MutS, L proteins inhibit RecA-catalyzed heteroduplex formation between M13 and fd DNA, but their potential influence on other steps in recombination has not been determined. Hence, we wanted to examine the role of mismatch repair when heterologous exchange is driven by RuvAB complex. The experiments described here show that RuvAB facilitate DNA exchange catalyzed by RecA protein when sequence heterology is encountered. The rate of open circular heteroduplex formation between M13-fd increased 2- 3-fold in the presence of these activities. This enhancement required the simultaneous presence of RuvA and RuvB proteins. Due to the levels of RecA used in this study there was no marked stimulation of M13-M13 exchange in the presence of RuvAB. Interestingly, in the presence of MutS, L, RuvAB failed to stimulate extensive strand exchange with <3% of linear substrate being converted to product. Substituting an ATPase defective mutant (MutS501) for wild-type reestablished the rate of open circular heteroduplex formation to that of RecA & RuvAB alone. Examination of reaction intermediates showed that MutS, L also blocked RuvAB-dependent branch migration in the absence of RecA. Taken together these results show that MutS, L modulate both RecA- and RuvAB-dependent branch migration between diverged DNA by distinct mechanisms. Consistent with replication fidelity this block by MR appears to be in response to mismatches and ATP-dependent DNA translocation by the MutS complex.